Rheumatic Associations of Autoimmune Thyroid Disease: a Systematic Review

Clinical Rheumatology (2019) 38:1801–1809 https://doi.org/10.1007/s10067-019-04498-1

REVIEW ARTICLE

Rheumatic associations of autoimmune thyroid disease: a systematic review

Clement E. Tagoe1,2 & Tejas Sheth3 & Eugeniya Golub1 & Karen Sorensen4

Received: 5 February 2019 /Revised: 22 February 2019 /Accepted: 27 February 2019 /Published online: 29 March 2019 # International League of Associations for Rheumatology (ILAR) 2019

Abstract To investigate specific disease patterns in the rheumatic manifestations associated with autoimmune thyroid disease (AITD) through a systematic literature review. We performed a systematic review using the Medline OVID, PubMed, EMBASE, and Web of Science databases through May 2018 for experimental and observational studies that explored the association of AITD with degenerative joint disease (DJD), osteoarthritis (OA), chronic widespread pain (CWP) and fibromyalgia syndrome (FMS), and seronegative inflammatory arthritis (IA). A total of 2132 articles were identified. After title and abstract screening and removal of duplicates, 66 articles were retrieved for full text review. Eighteen studies were deemed eligible for inclusion. Six observational studies reported up to 45% prevalence of DJD in AITD. Hand and spinal DJD were reportedly associated with higher odds of AITD. Twelve observational studies were retrieved reporting up to 62% prevalence of FMS in AITD patients. Four studies described the occurrence of seronegative IA in AITD patients. The rheumatic associations of AITD may manifest specific patterns of disease distinct from those of other well-defined autoimmune syndromes and contribute significantly to disease burden.

Keywords Autoimmunethyroiddisease .Chronicwidespreadpain .Fibromyalgia .Hashimotothyroiditis .Osteoarthritis .Spinal degenerative disc disease

Introduction proportion of patients with Graves’ disease may present with clinically significant hyperthyroidism and Graves’ The autoimmune thyroid diseases (AITD) comprise a spec- ophthalmopathy. Chronic lymphocytic thyroiditis (CLT) and trum of diseases with considerable clinical overlap that are its goitrous form Hashimoto thyroiditis (HT) are characterized typified histologically by the invasion of the thyroid gland by circulating antithyroid peroxidase antibodies (TPOAb) by lymphocytes [1]. Graves’ disease is characterized by the and/or antithyroglobulin antibodies (TgAb). Hashimoto thy- presence of antithyrotropin (TSH) receptor antibodies, respon- roiditis is the commonest cause of hypothyroidism, although sible for over activation of the thyroid gland. A significant at presentation patients can be euthyroid or even hyperthyroid [2]. Less well recognized are the rheumatic manifestations of AITD. The genetic associations of AITD overlap with those of Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10067-019-04498-1) contains supplementary the autoimmune connective tissue diseases (ACTD), and thus, material, which is available to authorized users. AITD is enriched in families with ACTD [3–7]. Indeed, AITD is the most common autoimmune disease with the prevalence * Clement E. Tagoe of CLT alone estimated conservatively to be around 10% of [email protected]; [email protected] the population in studies using the antithyroid autoantibodies TPOAb and TgAb as markers of disease [8]. Similarly, pa- 1 Division of Rheumatology, Department of Medicine, Albert Einstein tients with AITD have an increased incidence of ACTD, College of Medicine, Bronx, NY, USA which impacts the clinical presentation of the former [9]. 2 Division of Rheumatology, Montefiore Medical Center, 111 East Thyroid autoimmunity has been described as a classic ex- 210th Street, Bronx, NY 10467-2490, USA ample of organ-specific autoimmunity [10]. However, a grow- 3 Waterbury Hospital, Waterbury, CT, USA ing body of evidence suggests that thyroid autoimmunity per- 4 Albert Einstein College of Medicine, Bronx, NY, USA haps occurs as part of a more generalized autoimmune process 1802 Clin Rheumatol (2019) 38:1801–1809 in a significant proportion of patients where an overlap with Fibromyalgia and chronic widespread pain were defined using other autoimmune diseases exists [11, 12]. Furthermore, there the American College of Rheumatology (ACR) 1990 defini- is emerging evidence that rheumatic manifestations can occur tion and the updated criteria respectively [16, 17]. in AITD patients in the absence of well-defined ACTD or Osteoarthritis of the hip, knee, and hand were defined using clinically significant endocrine disease. Thus, association of the ACR criteria for each respectively [18–20]. Spinal osteo- AITD with chronic widespread pain (CWP) and fibromyalgia arthritis was defined by the radiographic presence of degener- syndrome (FMS), peripheral osteoarthritis (OA), spinal oste- ative joint disease (DJD) or degenerative disc disease (DDD) oarthritis, and seronegative inflammatory arthritis (IA) has respectively. Inflammatory arthritis was included if it was se- been described even in the absence of clinical hypothyroidism ronegative and not due to a well-defined ACTD, undifferen- [13]. Most of the emerging literature focuses on the rheumatic tiated connective tissue disease (UCTD), or other well-defined associations of CLT and HT which are much more common autoimmune syndrome and could thus be attributed to the than those of Graves’ disease. Despite the increasing appreci- presence of AITD alone. ation of the rheumatologic impact of AITD, the rheumatology The review was performed according to the Preferred community has been slow to incorporate testing for AITD into Reporting Items for Systematic Reviews and Meta-analyses clinical practice. (PRISMA) checklist ( S1 Table) and statement [21]. General reviews have suggested patterns of musculoskele- Experimental, quasi-experimental, and observational human tal manifestations of AITD, including OA, FMS and CWP, studies published in English which addressed the presence and spinal osteoarthritis. The objective of this review is to of the abovementioned comorbidities in patients with AITD provide an initial systematic review to investigate the avail- or vice versa were included. Qualitative research designs, re- able scientific evidence for a characteristic pattern of rheumat- view articles, meeting abstracts, protocols, and case reports ic manifestations of CLT and HT, outside of the musculoskel- were excluded. If a study addressed a heterogeneous group etal manifestations of overlap with connective tissue diseases of patients, the study was included if the results for patients or other autoimmune inflammatory syndromes. meeting our inclusion criteria were reported separately. The authors evaluated titles and abstracts of the references identified by means of the search strategy independently. The Materials and methods full texts of articles of interest were evaluated for inclusion. Any discrepancies between the reviewers were resolved by Search strategy consensus. Data pertaining to included studies were extracted independently using a standardized format by two authors (TS A comprehensive search was conducted in the MEDLINE and CT or EG and CT) and verified subsequently by the third OVID database as suggested by Haynes et al. (up until author. A narrative synthesis with tabulated results was per- May 31, 2018) [14, 15]. A medical librarian (KS) performed formed. Due to the high heterogeneity in design and outcome searches in three additional databases: PubMed; Embase on measures of the included studies, it was not feasible to com- Embase.com, which includes the years 1971–present; and bine data by means of meta-analysis. The study selection Web of Science, which includes the years 1985–present. For flowchart and results of the search strategy are shown in the PubMed searches, MeSH terms including Fig. 1. (BFibromyalgia,^ BChronic Pain,^ BOsteoarthritis,^ or BArthritis, Rheumatoid^)werecombinedwith(BThyroiditis, Autoimmune^ OR ((BThyroid Gland,^ OR BThyroid Results Diseases^)AND(BAutoimmune Diseases^ OR BAutoantibodies^[Mesh] OR BAutoimmunity^))). Title and Search outcome and level of evidence abstract words were included in the search. For the Embase search, Emtree terms and title and abstract words were A total of 2132 references were identified. After exclusion of searched. A topic search was conducted in Web of Science. duplicates, 1280 titles and abstracts were reviewed for rele- Supplementary Appendix A provides the complete PubMed vance to the topic. Sixty-six articles were reviewed in full text search. The associated rheumatic diseases were grouped into version. A total of 18 studies met the inclusion criteria for three groups: Group 1—degenerative arthritis or OA review, isolating studies that analyzed the MSK findings in (Supplementary Appendix A1); Group 2—FMS or CWP syn- AITD patients, not attributable to overlapping ACTD. Of the dromes (Supplementary Appendix A2); and Group 18, six, twelve, and four studies addressed the association of 3—seronegative IA (Supplementary Appendix A3). These AITD with osteoarthritis or degenerative joint disease, FMS searches were then complemented with a review of the refer- or CWP syndrome, and seronegative inflammatory arthritis, ence bibliographies of the review articles identified thus far to respectively. All the isolated studies were observational stud- detect any relevant studies left out by the initial search. ies. Level of evidence was judged using the grading of Clin Rheumatol (2019) 38:1801–1809 1803

Web of MEDLINE EMBASE PubMed namely peripheral and axial, of which 45% had spinal DJD. Science OA n = 18 OA n = 42 OA n = 14 OA n = 6 However, the study lacked a control arm [27]. Shrestha et al. FMS n = 17 FMS n =42 FMS n = 18 FMS n = 10 conducted a study specifically exploring the possible associa- IA n = 314 IA n = 755 IA n = 511 IA n = 385 tion between spinal DDD and AITD. The authors studied 4383 patients with radiographic evidence of spinal DDD and available data on the antithyroid autoantibodies TPOAb and TgAb, and reported a positive association between AITD and Combined result of the search the presence of spinal DJD with an adjusted OR of 1.8 (95% (n = 2132) CI 1.6–2.2) [28]. Duplicates (n = 852) Fibromyalgia and chronic widespread pain Titles and abstracts reviewed Excluded – (n = 1214) (n = 1280) Abstracts = 283 in autoimmune thyroid disease Books/Book secons = 13 Improperly referenced = 18 A total of 12 observational studies addressing the impact of Non-rheumatologic aspects Full text reviewed of AITD & Reviews = 900 FMS and CWP in patients with AITD or vice versa were (n = 66) identified by the search (Table 2). Excluded with reasons (n = 48) In the study by Soy and colleagues mentioned in associa- Well-deﬁned ACTD tion with OA, the authors also examined FMS in patients with associaons = 19 Non-rheumac disease AITD [23]. They found FMS to be the most common rheu- Studies included associaons = 4 matic association in the 65 patients with AITD examined, with (n = 18) Not meeng inclusion a prevalence rate of 31% as defined by the 1990 ACR criteria OA = 6, FMS = 12, IA = 4 criteria = 25 [16]. The referral base was an endocrinology clinic and only Fig. 1 Flow diagram of the search methodology 5% of patients were euthyroid; almost half were hypothyroid. Bazzichi et al. examined 120 FMS patients and found that the prevalence of AITD was 41%. The actual prevalence in the recommendation, assessment, development, and evaluation general FMS population could be higher since the authors (GRADE) approach [7]. Risk of bias was assessed using the reported that basal thyroid hormone levels of their FMS pa- tool for nonrandomized studies by Kim et al. [22]. tients were in the normal range [29]. In a follow-up study, Bazzichi and coworkers were again able to show an associa- Osteoarthritis in autoimmune thyroid disease tion of AITD with FMS of about 30%, which did not exist in patients with subclinical hypothyroidism in the absence of Six observational studies addressing the impact of OA in pa- AITD. They concluded that the presence of AITD was likely tients with AITD or vice versa were identified by the search important in the development of FMS and deserved further and met the inclusion criteria for review (Table 1). Most of study [30]. Aarflot and colleagues showed a significant asso- these studies were cross-sectional; two studies explored pos- ciation between CWP and thyroid autoimmunity in a commu- sible association between spinal OA and hand OA, and AITD. nity survey of 737 men and 771 women [31]. The prevalence Studies used the terms DJD and DDD interchangeably in ref- was significantly higher in persons with CWP than without erence to axial OA. (16.0% versus 7.39%, p < 0.01). Tagoe et al. specifically Soy et al. while investigating the frequency of rheumatic looked at 46 AITD patients with normal thyroid function diseases among 65 patients with AITD found symptomatic and the absence of overlap with well-defined connective tissue OA as the third most common comorbid condition (15%) in disease and reported a 56% prevalence of FMS defined by this population, behind FMS and recurrent aphthous stomatitis 1990 ACR criteria. This prevalence was higher than in previ- [23]. It manifested mainly as hand OA (Heberden’snodes) ous studies, perhaps because the study was performed at a and knee OA. Mobini et al. described the presence of AITD tertiary care center with the possibility of referral bias [27]. in 36.3% of 80 OA patients [24]. Addimanda and coworkers Other studies that have examined the prevalence of in their investigation of the clinical associations of hand OA AITD in the FMS population include the work of Ribeiro studied 446 patients and found that they had significantly et al. who reported a cross-sectional study of 146 patients increased odds of having AITD (odd’s ratio (OR) 4.85, 95% with FMS defined by 1990 ACR criteria. Patients were CI 1.77–13.29) compared to 307 controls [25]. Hezarkhani evaluated for the presence of either TPOAb and/or TgAb. et al. found a prevalence rate for OA of 25% in their cohort The authors reported an association between AITD and which was 86% female with a mean age of 39 years [26]. In a FMS with an OR of 3.87 (95% CI 1.54–10.13) [32]. recent study, Tagoe et al. reported that 88% of their cohort of Pamuk et al. detected a 34.4% frequency of at least one 46 AITD patients had imaging evidence of OA at all sites, thyroid antibody in a cohort of 128 patients with FMS [33]. 1804 Clin Rheumatol (2019) 38:1801–1809

Table 1 Characteristics of observational studies for OA and AITD

Study year Study author Country and source Study population Outcome and results Level of Risk of population evidence bias

2007 Soy et al. [23] Cross-sectional study at an 65 patients with AITD 15% prevalence of OA Low Low endocrinology clinic in No control arm Turkey 2011 Mobini et al. [24] Cross-sectional study in 80 patients with RA 36.3% prevalence of AITD as Very low High Pakistan compared to 80 patients defined by presence of with OA TPOAb 2012 Addimanda et al. [25] Prospective study at 3 446 patients with hand OA 5.4% prevalence of AITD as Low High tertiary rheumatology 307 control subjects defined by presence of clinics in Italy TPOAb and/or TgAb 2013 Tagoe et al. [27] Retrospective cohort study 65 patients with AITD 88% prevalence of any OA Low Low at a tertiary No control arm findings rheumatology clinic in 45% prevalence of spinal USA DDD 2014 Hezarkhani et al. [26] Cross-sectional study at a 65 patients with AITD 23% prevalence of OA Low Low rheumatology clinic in No control arm Iran 2016 Shrestha et al. [28] Cross-sectional study at a 1557 patients with AITD AITD is associated with Low Low tertiary rheumatology No control arm higher frequency of spinal clinic in USA DDD with an adjusted OR of 1.8

In a recent study, Suk et al. reported similar results to the cohort, 10.7% had Raynaud’s phenomenon, and 5.3% had other studies and showed that among patients with normal fibromyalgia compared to the Turkish female population thyroid-stimulating hormone (TSH) levels (mean TSH FMS prevalence rate of 3.6%. The sample size was small 1.58 mIU/L), specifically in a cohort of 149 FMS patients and lacked a control arm making generalization of findings and 68 healthy controls, there was a 19% prevalence of difficult [26]. Baskan et al. studied the relationship between anti-TPO antibodies in the FMS patients compared to 7% FMS, AITD, and RA by examining the prevalence of TPOAb in the controls, which was statistically significant [34]. and TgAb in 65 FMS, 39 RA, and 40 control patients respec- Two studies have attempted to evaluate the presence of tively. Although a trend towards high TPOAb and TgAb was thyroid autoimmunity and/or FMS in the specific context of seen in FMS, the sample size was small and did not reach rheumatoid arthritis (RA). Pamuk et al. examined thyroid au- statistical significance [36]. toimmunity in 128 euthyroid FMS patients, 64 RA patients, Using a population of FMS patients in whom the updated and 64 healthy control subjects [33]. Thyroid autoimmunity ACR classification criteria were used, Haliloglu and col- was found in 34.4% of FMS patients and 29.7% of RA pa- leagues found a prevalence rate of 62% for fibromyalgia pain tients, significantly higher than in controls (18.8%) (p <0.05). in 79 consecutive patients with HT, and they noted the higher Furthermore, 20.3% of FMS patients had positive TgAb and prevalence compared to the rate of 30 to 40% using the older 24.2% had positive TPOAb. The differences were significant classification criteria. There was also a strong positive corre- for TPOAb between FMS and RA as compared to controls. lation between the presence of FMS and being positive for Ahmad et al. reported a similar frequency of TPOAb positiv- TPOAb [37, 38]. ity (29%) in a cohort of 203 RA patients [35]. The authors reported a 40% prevalence of FMS or CWP in TPOAb and/or TgAb-positive patients versus 17% for thyroid antibody- Inflammatory arthritis in autoimmune thyroid disease negative RA patients. Logistic regression analyses adjusted by age, sex, diabetes, and BMI indicated significantly higher In a survey-based study, Mosca et al. found a 34% prevalence odds (OR = 4.64, 95% CI = 2.11–10.20) of development of of AITD among the first and second-degree relatives of 626 FMS in RA patients with TPOAb-positive thyroid patients with inflammatory arthritis [39]. In another study, Soy autoimmunity. et al. reported a 6% prevalence of inflammatory arthritis In the same brief report mentioned above in reference to among the 65 patients with documented AITD mentioned OA, Hezarkhani and coworkers described the rheumatic man- earlier. However, according to the authors, the patients had ifestations of a cohort of 65 AITD patients; 56 of whom were arthritis patterns consistent with RA, psoriatic arthritis (PsA), women. They found that 39.3% of them had carpal tunnel and arthritis associated with mixed connective tissue disease syndrome, the commonest rheumatic presentation in their (MCTD) [23]. The search identified four studies that lnRemtl(09 38:1801 (2019) Rheumatol Clin

Table 2 Characteristics of observational studies for FMS and AITD

Study year Study author Country and source population Study population Outcome and results Level of evidence Risk of bias

1996 Aarflot et al. [31] Cross-sectional study in Norway 1508 volunteers attending Higher prevalence of thyroid autoimmunity in Low Low – 1809 National Health Screening persons with than without CWP (16.0% Service’s mobile unit vs 7.3%, p < 0.01) 2004 Riberio el al. [32] Cross-sectional study at a 146 women with FMS Association between FMS and AITD after Low High rheumatology clinic in Brazil 76 control patients adjustment for depression and age, OR = 4.52 2007 Pamuk et al. [33] Prospective study at a rheumatology 128 euthyroid FMS patients 34.4% prevalence of AITD in FMS Low Low clinic in Turkey 64 RA patients population, significantly higher than 64 healthy controls controls (18.8%, p < 0.05) 2007 Soy et al. [23] Cross-sectional study at an 65 patients with AITD FMS most frequent rheumatic disease in Low Low endocrinology clinic in Turkey No control arm AITD patients, 31% prevalence 2007 Bazzichi et al. [29] Prospective study at a rheumatology 120 FMS patients Higher frequency of AITD in post-menopausal Low High clinic in Italy No control arm patients as compared to pre-menopausal patients 2010 Baskan et al. [36] Cross-sectional study in Turkey 65 FMS patients 15.4% prevalence of AITD in FMS patients, Low High 39 RA patients no statistical significant differences in rates 40 healthy controls of TPOAb and TgAb positivity among the groups 2011 Erkoc et al. [64] Prospective study 30 patients with AITD and 50% prevalence of FMS Low High thyroid dysfunction 2012 Bazzichi et al. [30] Prospective study at a rheumatology 52 patients with AITD 31% prevalence of FMS Low Low clinic in Italy 25 healthy subjects 2012 Suk et al. [34] Prospective study at a rheumatology 149 euthyroid FMS patients Euthyroid FMS patients with significantly Low Low clinic in the Republic of Korea 68 healthy controls higher prevalence of AITD as compared to age and sex-matched control (19% vs 7%, p 0.04) 2013 Tagoe et al. [27] Retrospective cohort study at a tertiary 65 patients with AITD 59% prevalence of FMS Low Low rheumatology clinic in the USA No control arm 2014 Hezarkhani et al. [26] Cross-sectional study at a rheumatology 65 patients with AITD 5.3% prevalence of FMS Low Low clinic in Iran No control arm 2017 Haliloglu et al. [37] Cross-sectional study in Turkey 79 consecutive patients with HT Higher prevalence of FMS than using old Low Low using updated classification classification criteria (62%) criteria for FMS 1805 1806 Clin Rheumatol (2019) 38:1801–1809

Table 3 Characteristics of observational studies for IA and AITD

Study year Study author Country and source Study population Outcome and results Level of evidence Risk of bias population

1984 LeRiche et al. [40] Prospective study at a 15 patients with AITD 60% prevalence of Low Low rheumatology clinic in No control arm seronegative IA Canada 2013 Tagoe et al. [27] Retrospective cohort study 65 patients with AITD 26% prevalence of Low Low at a tertiary rheumatology No control arm seronegative IA clinic in the USA 1997 Punzi et al. [42] Prospective study with mean 33 patients with CLT 25% prevalence of High Low follow-up of 6.4 years in Italy and arthritis seronegative IA 1993 Golding [41] Observational study 11 patients with AITD No prevalence data Very low High and arthritis described seronegative IA solely attributable to AITD in var- dysfunction and had a clinical pattern similar to the arthritis ious cohorts (Table 3). found in the context of connective tissue diseases [42]. The Lerich and Bell described in 1984 a group of 15 patients small number of study subjects did not allow for with HT who presented with inflammatory polyarthritis. Two generalization. were hypothyroid, five were euthyroid, and eight had incipient hypothyroidism, defined by elevated TSH or mild symptoms. Nine patients had non-erosive seronegative polyarthritis Discussion which the authors felt was a unique presentation of HT while the rest had rheumatoid factor positive erosive arthritis indis- We have performed a systematic review of the rheumatic man- tinguishable from seropositive rheumatoid arthritis [40]. ifestations of AITD with specific reference to CWP and FMS, Similarly, Golding described 11 patients with AITD who peripheral osteoarthritis, axial OA, and IA. The quality of had high levels of antithyroid antibodies and were euthyroid evidence was deemed to be low for most of the studies. or marginally hypothyroid. These patients had clinical fea- However, using the GRADE protocol, several of the studies tures of a seronegative inflammatory polyarthritis resembling could be upgraded to a level of moderate based on the odds early, mild rheumatoid disease. In these patients, the joint ratios of association exceeding two [43]. These findings sug- changes markedly improved with thyroid therapy, whereas gest that AITD may be a risk factor for CWP and FMS, pe- in the patients with more definite rheumatoid-like features ripheral osteoarthritis, and axial OA. The association with IA and positive RF, the joint symptoms were unaffected by thy- was less definite based on the available data. Elatter et al. roid therapy. Thus, it was believed that this was a group of studied 150 RA patients and found a 24% prevalence of hy- patients with a distinctive non-rheumatoid syndrome of sero- pothyroidism [44]. High rates of association between RA and negative polyarthritis associated with AITD. This study AITD are reported in other studies, some of which report an lacked a control arm [41]. influence on disease prognosis [45–47]. However, these re- In a more recent report, Tagoe et al. found that 26% of a ports did not suggest AITD as a cause of inflammatory arthri- cohort of 46 euthyroid patients with no known well-defined tis and the possible influence of AITD on RA disease severity connective tissue disease or spondyloarthropathy (SpA) had was not a focus of this review. Insufficient data were available evidence of synovitis in the wrists and hands. Two patients for other reported rheumatic disease associations like adhesive (4%) had seronegative arthritis resembling RA with evidence capsulitis, carpal tunnel syndrome, Dupuytren’scontracture, of spinal arthritis and FMS. The authors suggested a separate and trigger finger, to be included in this review [48]. etiology for the inflammatory arthritis, perhaps related to The association between AITD and osteoarthritis has been AITD [27]. known for at least a century [49]. Early reports focused spe- Punzi et al. conducted a prospective study to explore the cifically on the association with hypothyroidism. Subsequent natural history of arthritis in patients with AITD [42]. During a publications including one by Gillan and colleagues describ- mean follow-up of 6.4 years of 33 patients with CLT and ing knee OA were able to clearly identify the association with arthritis, almost half of the patients with polyarthritis devel- AITD [50]. Further studies have extended findings of associ- oped severe RA characterized by bony erosions, high levels of ation to include spinal degeneration and axial OA [28]. IL-1, and increased frequency of HLA DR4. The other half Fibromyalgia and CWP account for an estimated 10% and had seronegative, non-erosive arthritis with a milder clinical 30% respectively of outpatient rheumatology visits [51, 52]. course, low levels of IL-1, and increased frequency of HLA Thus, a significant association between AITD and FMS/CWP DR3. The arthritis was independent of the degree of thyroid would suggest a significant burden of disease attributable to Clin Rheumatol (2019) 38:1801–1809 1807 the former. Like the connection between AITD and OA, a on observational reports and the possibility that some patients possible link with generalized pain is not a new finding. had unrecognized well-differentiated connective tissue or au- Becker et al. described in 1963 a significant association be- toimmune disease. Secondly, given the extended study period tween HT and fibrositis, a term now supplanted by FMS, in 40 and the lack of uniformity between studies, there is likely to (7.9%) of 506 patients [53]. Our current understanding of the have been a lack of uniformity of diagnostic procedures be- association of AITD with FMS is now well documented and tween studies. This heterogeneity precluded the performance hasbeenreviewedelsewhere[54]. of a meta-analysis of the data. Large epidemiologic studies of Although the literature supports the association between high quality are required to confirm the specificity of these AITD and rheumatic manifestations, the precise role of patterns of musculoskeletal associations with AITD. AITD in terms of direct causation versus modulation of the Recognizing the association is crucial to understanding any rheumatic presentation of co-existing autoimmune disease impact AITD may have on disease burden alone or in combi- cannot be determined. Some evidence for AITD influencing nation with well-defined ACTD. the disease severity of some ACTD exists, and as noted be- We believe the preponderance of the evidence available fore, thyroid autoimmunity occurs with enhanced frequency now justifies the assessment of thyroid autoantibodies as part in patients with ACTD [55–57]. of routine rheumatology practice to better assess the disease Franco et al. looked at AITD in Colombian patients with burden of AITD in rheumatic diseases, not only generally SLE and concluded that the former is more frequent in lupus through its endocrine and cardiovascular effects, but specifi- but does not influence disease severity [58]. Appenzeller et al. cally due to its rheumatic manifestations. The association of examined the presence of AITD in 524 SLE patients and endocrine diseases like diabetes with musculoskeletal disease found thyroid autoantibodies in 17% who were euthyroid. is well recognized [63]. The association of AITD with rheu- About 6% of the cohort had hypothyroidism and 11.5% had matic syndromes may involve metabolic processes as well as subclinical hypothyroidism. Overall, about 70% of those with immunologic and inflammatory pathways. This review sug- thyroid function abnormalities had antithyroid autoantibodies. gests that AITD, which has a population prevalence of about In this study, SLE disease activity was positively correlated 10 to 20%, may indeed have an association with osteoarthritis with occurrence of symptoms of hyperthyroidism [59]. These and chronic widespread pain, implicating it as an important studies suggest that AITD may influence the severity of other risk factor for these conditions. A detailed study of this asso- autoimmune syndromes that it overlaps with although the di- ciation might help to provide a better understanding of these rection of that influence is yet to be defined. The current study syndromes which could facilitate a more descriptive overview was not designed to address the non-rheumatologic impact of and allow a departure from our normative classifications of AITD on disease burden for example through effects on insu- OA and chronic widespread pain. lin resistance, metabolic syndrome, and cardiovascular risk which have been reported in the literature [60–62]. Compliance with ethical standards

Disclosures None. Conclusion

The presence of AITD not only increases the risk of clinically significant thyroid disease such as hypothyroidism in the par- ticular case of CLTand HT, but possibly increases the frequen- References cy of axial OA, and probably increases the presence of peripheral OA characterized by Heberden’s and Bouchard’snodes, 1. Dayan CM, Daniels GH (1996) Chronic autoimmune thyroiditis. N knee, and hip OA. It probably increases the prevalence of Engl J Med 335(2):99–107. https://doi.org/10.1056/ FMS in some populations, including AITD with no ACTD nejm199607113350206 overlap and in RA patients [35]. The quality of the evidence 2. Caturegli P, De Remigis A, Rose NR (2014) Hashimoto thyroiditis: is low to moderate at best. However, the high strength of the clinical and diagnostic criteria. Autoimmun Rev 13(4–5):391–397. https://doi.org/10.1016/j.autrev.2014.01.007 associations described particularly for peripheral OA and 3. Weetman AP (2009) The genetics of autoimmune thyroid disease. CWP/FMS call for further studies particularly given the very Horm Metab Res 41(6):421–425. https://doi.org/10.1055/s-0029- high prevalence of AITD in the general population and its 1214415 enrichment in ACTD. Furthermore, the data suggest a specific 4. Cho JH, Gregersen PK (2011) Genomics and the multifactorial pattern of association involving peripheral arthritis including nature of human autoimmune disease. N Engl J Med 365(17): 1612–1623. https://doi.org/10.1056/NEJMra1100030 Heberden’sandBouchard’s nodes, spinal involvement of the 5. Tomer Y (2014) Mechanisms of autoimmune thyroid diseases: cervical and lumbar discs, and chronic widespread pain or from genetics to epigenetics. Annu Rev Pathol 9:147–156. https:// fibromyalgia. A limitation of the study was its dependence doi.org/10.1146/annurev-pathol-012513-104713 1808 Clin Rheumatol (2019) 38:1801–1809

6. Farh KK, Marson A, Zhu J, Kleinewietfeld M, Housley WJ, Beik S, and Therapeutic Criteria Committee of the American Rheumatism Shoresh N, Whitton H, Ryan RJ, Shishkin AA, Hatan M, Carrasco- Association. Arthritis Rheum 29(8):1039–1049 Alfonso MJ, Mayer D, Luckey CJ, Patsopoulos NA, De Jager PL, 20. Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D, Kuchroo VK, Epstein CB, Daly MJ, Hafler DA, Bernstein BE Brandt K, Brown C, Cooke TD, Daniel W, Gray R, Greenwald R, (2015) Genetic and epigenetic fine mapping of causal autoimmune Hochberg M, Howell D, Ike R, Kapila P, Kaplan D, Koopman W, disease variants. Nature 518(7539):337–343. https://doi.org/10. Longley S, Mcshane DJ, Medsger T, Michel B, Murphy W, Osial T, 1038/nature13835 Ramsey-Goldman R, Rothschild B, Stark K, Wolfe F (1990) The 7. Guyatt GH, Oxman AD, Kunz R, Vist GE, Falck-Ytter Y, American College of Rheumatology criteria for the classification Schunemann HJ (2008) What is Bquality of evidence^ and why is and reporting of osteoarthritis of the hand. Arthritis Rheum 33(11): it important to clinicians? BMJ 336(7651):995–998. https://doi. 1601–1610 org/10.1136/bmj.39490.551019.BE 21. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred 8. Hollowell JG, Staehling NW, Flanders WD, Hannon WH, Gunter reporting items for systematic reviews and meta-analyses: the EW, Spencer CA, Braverman LE (2002) Serum TSH, T(4), and PRISMA statement. BMJ 339:b2535 thyroid antibodies in the United States population (1988 to 1994): 22. Kim SY,Park JE, Lee YJ, Seo HJ, Sheen SS, Hahn S, Jang BH, Son National Health and Nutrition Examination Survey (NHANES III). HJ (2013) Testing a tool for assessing the risk of bias for J Clin Endocrinol Metab 87(2):489–499 nonrandomized studies showed moderate reliability and promising 9. Somers EC, Thomas SL, Smeeth L, Hall AJ (2006) Autoimmune validity. J Clin Epidemiol 66(4):408–414. https://doi.org/10.1016/j. diseases co-occurring within individuals and within families: a sys- jclinepi.2012.09.016 – tematic review. Epidemiology 17(2):202 217. https://doi.org/10. 23. Soy M, Guldiken S, Arikan E, Altun BU, Tugrul A (2007) 1097/01.ede.0000193605.93416.df Frequency of rheumatic diseases in patients with autoimmune thy- 10. Rapoport B, McLachlan SM (2001) Thyroid autoimmunity. J Clin roid disease. Rheumatol Int 27(6):575–577. https://doi.org/10. – Invest 108(9):1253 1259. https://doi.org/10.1172/JCI14321 1007/s00296-006-0263-8 11. Boelaert K, Newby PR, Simmonds MJ, Holder RL, Carr-Smith JD, 24. Mobini M, Kashi Z, Ravanbakhsh N (2011) Thyroid disorders in Heward JM, Manji N, Allahabadia A, Armitage M, Chatterjee KV, rheumatoid arthritis and osteoarthritis. Pak J Med Sci 27(3):595–598 Lazarus JH, Pearce SH, Vaidya B, Gough SC, Franklyn JA (2010) 25. Addimanda O, Mancarella L, Dolzani P, Ramonda R, Fioravanti A, Prevalence and relative risk of other autoimmune diseases in subjects Brusi V, Pignotti E, Meliconi R (2012) Clinical associations in – with autoimmune thyroid disease. Am J Med 123(2):183.e181 patients with hand osteoarthritis. Scand J Rheumatol 41(4):310– 183.e189. https://doi.org/10.1016/j.amjmed.2009.06.030 313. https://doi.org/10.3109/03009742.2012.656699 12. Biro E, Szekanecz Z, Czirjak L, Danko K, Kiss E, Szabo NA, Szucs 26. Hezarkhani S, Sedighi S, Aghaei M, Shamekhi M, Nomali M G, Zeher M, Bodolay E, Szegedi G, Bako G (2006) Association of (2014) Rheumatologic manifestations in Iranian patients with auto- systemic and thyroid autoimmune diseases. Clin Rheumatol 25(2): immune thyroid diseases. J Clin Diagn Res 8(10):MC06–MC08. 240–245. https://doi.org/10.1007/s10067-005-1165-y https://doi.org/10.7860/JCDR/2014/7974.4981 13. Tagoe CE, Zezon A, Khattri S (2012) Rheumatic manifestations of 27. Tagoe CE, Zezon A, Khattri S, Castellanos P (2013) Rheumatic autoimmune thyroid disease: the other autoimmune disease. J manifestations of euthyroid, anti-thyroid antibody-positive patients. Rheumatol 39(6):1125–1129. https://doi.org/10.3899/jrheum. Rheumatol Int 33(7):1745–1752. https://doi.org/10.1007/s00296- 120022 012-2616-9 14. Haynes RB, Wilczynski NL (2004) Optimal search strategies for 28. Shrestha A, Cohen HW, Tagoe CE (2016) Association of spinal retrieving scientifically strong studies of diagnosis from Medline: degenerative disc disease with thyroid autoimmunity. Clin Exp analytical survey. BMJ 328(7447):1040. https://doi.org/10.1136/ Rheumatol 33(7):1745–1752. https://doi.org/10.1007/s00296-012- bmj.38068.557998.EE 2616-9 15. Haynes RB, Wilczynski N, McKibbon KA, Walker CJ, Sinclair JC (1994) Developing optimal search strategies for detecting clinically 29. Bazzichi L, Rossi A, Giuliano T, De Feo F, Giacomelli C, Consensi sound studies in MEDLINE. J Am Med Inform Assoc 1(6):447–458 A, Ciapparelli A, Consoli G, Dell'osso L, Bombardieri S (2007) 16. Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Association between thyroid autoimmunity and fibromyalgic dis- – Goldenberg DL, Tugwell P, Campbell SM, Abeles M, Clark P, ease severity. Clin Rheumatol 26(12):2115 2120. https://doi.org/ Fam AG, Farber SJ, Fiechtner JJ, Michael Franklin C, Gatter RA, 10.1007/s10067-007-0636-8 Hamaty D, Lessard J, Lichtbroun AS, Masi AT, Mccain GA, John 30. Bazzichi L, Rossi A, Zirafa C, Monzani F, Tognini S, Dardano A, Reynolds W, Romano TJ, Jon Russell I, Sheon RP (1990) The Santini F, Tonacchera M, De Servi M, Giacomelli C, De Feo F, American College of Rheumatology 1990 Criteria for the Doveri M, Massimetti G, Bombardieri S (2012) Thyroid autoim- Classification of Fibromyalgia. Report of the Multicenter Criteria munity may represent a predisposition for the development of fi- – Committee. Arthritis Rheum 33(2):160–172 bromyalgia? Rheumatol Int 32(2):335 341. https://doi.org/10. 17. Wolfe F, Clauw DJ, Fitzcharles MA, Goldenberg DL, Katz RS, 1007/s00296-010-1620-1 Mease P, Russell AS, Russell IJ, Winfield JB, Yunus MB (2010) 31. Aarflot T, Bruusgaard D (1996) Association between chronic wide- The American College of Rheumatology preliminary diagnostic spread musculoskeletal complaints and thyroid autoimmunity. criteria for fibromyalgia and measurement of symptom severity. Results from a community survey. Scand J Prim Health Care – Arthritis Care Res (Hoboken) 62(5):600–610. https://doi.org/10. 14(2):111 115 1002/acr.20140 32. Ribeiro LS, Proietti FA (2004) Interrelations between fibromyalgia, 18. Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D, thyroid autoantibodies, and depression. J Rheumatol 31(10):2036–2040 Brandt K, Brown C, Cooke TD, Daniel W, Feldman D et al 33. Pamuk ON, Cakir N (2007) The frequency of thyroid antibodies in (1991) The American College of Rheumatology criteria for the fibromyalgia patients and their relationship with symptoms. Clin classification and reporting of osteoarthritis of the hip. Arthritis Rheumatol 26(1):55–59. https://doi.org/10.1007/s10067-006-0237-y Rheum 34(5):505–514 34. Suk JH, Lee JH, Kim JM (2012) Association between thyroid au- 19. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, toimmunity and fibromyalgia. Exp Clin Endocrinol Diabetes Christy W, Cooke TD, Greenwald R, Hochberg M et al (1986) 120(7):401–404. https://doi.org/10.1055/s-0032-1309008 Development of criteria for the classification and reporting of oste- 35. Ahmad J, Blumen H, Tagoe CE (2015) Association of antithyroid oarthritis. Classification of osteoarthritis of the knee. Diagnostic peroxidase antibody with fibromyalgia in rheumatoid arthritis. Clin Rheumatol (2019) 38:1801–1809 1809

Rheumatol Int 35(8):1415–1421. https://doi.org/10.1007/s00296- 51. Branco JC, Bannwarth B, Failde I, Abello Carbonell J, Blotman F, 015-3278-1 Spaeth M, Saraiva F, Nacci F, Thomas E, Caubere JP, Le Lay K, 36. Başkan BM, Sivas F, Aktekin LA, Yurdakul FG, Çinar NK, Bodur Taieb C, Matucci-Cerinic M (2010) Prevalence of fibromyalgia: a H, Özoran K (2010) Relationship between thyroid autoimmunity survey in five European countries. Semin Arthritis Rheum 39(6): and depression, quality of life, and disease symptoms in patients 448–453. https://doi.org/10.1016/j.semarthrit.2008.12.003 with fibromyalgia and rheumatoid arthritis. Turk J Rheumatol 52. Vanhoof J, Declerck K, Geusens P (2002) Prevalence of rheumatic 25(3):130–136 diseases in a rheumatological outpatient practice. Ann Rheum Dis 37. Haliloglu SEB, Uzkeser H, Sevimli H, Carlioglu A, Macit PM 61(5):453–455 (2017) Fibromyalgia in patients with thyroid autoimmunity: prev- 53. Becker KL, Ferguson RH, Mc CW (1963) The connective-tissue alence and relationship with disease activity. Clin Rheumatol 36(7): diseases and symptoms associated with Hashimoto’s thyroiditis. N 1617–1621. https://doi.org/10.1007/s10067-017-3556-2 Engl J Med 268:277–280. https://doi.org/10.1056/ 38. Wolfe F, Clauw DJ, Fitzcharles MA, Goldenberg DL, Hauser W, nejm196302072680601 Katz RS, Mease P, Russell AS, Russell IJ, Winfield JB (2011) 54. Ahmad J, Tagoe CE (2014) Fibromyalgia and chronic widespread Fibromyalgia criteria and severity scales for clinical and epidemio- pain in autoimmune thyroid disease. Clin Rheumatol 33(7):885– logical studies: a modification of the ACR Preliminary Diagnostic 891. https://doi.org/10.1007/s10067-014-2490-9 – Criteria for Fibromyalgia. J Rheumatol 38(6):1113 1122. https:// 55. Pyne D, Isenberg DA (2002) Autoimmune thyroid disease in sys- doi.org/10.3899/jrheum.100594 temic lupus erythematosus. Ann Rheum Dis 61(1):70–72 39. Mosca M, Carli L, d'Ascanio A, Tani C, Talarico R, Baldini C, 56. Pan XF, Gu JQ, Shan ZY (2015) Increased risk of thyroid autoim- BazzichiL,TavoniA,MiglioriniP,BombardieriS(2008) munity in rheumatoid arthritis: a systematic review and meta-anal- Occurrence of organ-specific and systemic autoimmune diseases ysis. Endocrine 50(1):79–86. https://doi.org/10.1007/s12020-015- among the first- and second-degree relatives of Caucasian patients 0533-x with connective tissue diseases: report of data obtained through direct patient interviews. Clin Rheumatol 27(8):1045–1048. 57. Pan XF, Gu JQ, Shan ZY (2015) Patients with systemic lupus https://doi.org/10.1007/s10067-008-0904-2 erythematosus have higher prevalence of thyroid autoantibodies: a 40. LeRiche NG, Bell DA (1984) Hashimoto's thyroiditis and systematic review and meta-analysis. PLoS One 10(4):e0123291. polyarthritis: a possible subset of seronegative polyarthritis. Ann https://doi.org/10.1371/journal.pone.0123291 Rheum Dis 43(4):594–598 58. Franco JS, Amaya-Amaya J, Molano-Gonzalez N, Caro-Moreno J, 41. Golding DN (1993) Rheumatism and the thyroid. J R Soc Med Rodriguez-Jimenez M, Acosta-Ampudia Y, Mantilla RD, Rojas- 86(3):130–132 Villarraga A, Anaya JM (2015) Autoimmune thyroid disease in 42. Punzi L, Michelotto M, Pianon M, Bertazzolo N, Fagiolo U, Colombian patients with systemic lupus erythematosus. Clin – Betterle C, Vettor R, Todesco S (1997) Clinical, laboratory and Endocrinol 83(6):943 950. https://doi.org/10.1111/cen.12662 immunogenetic aspects of arthritis associated with chronic lympho- 59. Appenzeller S, Pallone AT, Natalin RA, Costallat LT (2009) cytic thyroiditis. Clin Exp Rheumatol 15(4):373–380 Prevalence of thyroid dysfunction in systemic lupus erythematosus. 43. Atkins D, Best D, Briss PA, Eccles M, Falck-Ytter Y, Flottorp S, J Clin Rheumatol 15(3):117–119. https://doi.org/10.1097/RHU. Guyatt GH, Harbour RT, Haugh MC, Henry D, Hill S, Jaeschke R, 0b013e31819dbe4c Leng G, Liberati A, Magrini N, Mason J, Middleton P, Mrukowicz 60. Dessein PH, Joffe BI, Stanwix AE (2004) Subclinical hypothyroid- J, O'Connell D, Oxman AD, Phillips B, Schunemann HJ, Edejer T, ism is associated with insulin resistance in rheumatoid arthritis. Varonen H, Vist GE, Williams JW Jr, Zaza S (2004) Grading quality Thyroid 14(6):443– 446. https://doi.org/10.1089/ of evidence and strength of recommendations. BMJ 328(7454): 105072504323150750 1490. https://doi.org/10.1136/bmj.328.7454.1490 61. Zhuo Q, Yang W, Chen J, Wang Y (2012) Metabolic syndrome 44. Elattar EA, Younes TB, Mobasher SA (2014) Hypothyroidism in meets osteoarthritis. Nat Rev Rheumatol 8(12):729–737. https:// patients with rheumatoid arthritis and its relation to disease activity. doi.org/10.1038/nrrheum.2012.135 Egypt Rheumatol Rehabil 41(2):58–65 62. Raterman HG, van Halm VP, Voskuyl AE, Simsek S, Dijkmans 45. Joshi P, Agarwal A, Vyas S, Kumar R (2016) Prevalence of hypo- BA, Nurmohamed MT (2008) Rheumatoid arthritis is associated thyroidism in rheumatoid arthritis and its correlation with disease with a high prevalence of hypothyroidism that amplifies its cardio- activity. Trop Dr 47:6–10. https://doi.org/10.1177/ vascular risk. Ann Rheum Dis 67(2):229–232. https://doi.org/10. 0049475515627235 1136/ard.2006.068130 46. Raterman HG, Voskuyl AE, Simsek S, Schreurs MW, van 63. Smith LL, Burnet SP, McNeil JD (2003) Musculoskeletal manifes- Hoogstraten IM, Peters MJ, van Halm VP, Dijkmans BA, Lips P, tations of diabetes mellitus. Br J Sports Med 37(1):30–35 Lems WF, Nurmohamed MT (2013) Increased progression of ca- 64. Erkoç S, Karaahmet ÖZ, Bal A, Çakci A (2011) Evaluation of rotid intima media thickness in thyroid peroxidase antibodies- thyroid function in patients presenting with musculoskeletal com- positive rheumatoid arthritis patients. Eur J Endocrinol / plaints. J Rheumatol Med Rehabil 22(1):1–7 European Federation of Endocrine Societies 169(6):751–757. https://doi.org/10.1530/eje-13-0394 47. Koszarny A, Majdan M, Suszek D, Wielosz E, Dryglewska M (2013) Relationship between rheumatoid arthritis activity and anti- Publisher’snoteSpringer Nature remains neutral with regard to jurisdic- thyroid antibodies. Pol Arch Med Wewn 123(7–8):394–400 tional claims in published maps and institutional affiliations. 48. Cakir M, Samanci N, Balci N, Balci MK (2003) Musculoskeletal manifestations in patients with thyroid disease. Clin Endocrinol Protocol registration https://www.researchregistry.com/browse-the- 59(2):162–167. https://doi.org/10.1046/j.1365-2265.2003.01786.x registry#home/researchregistry4536 49. Doyle L (1991) Myxoedema: some early reports and contributions by British authors, 1873-1898. J R Soc Med 84(2):103–106 50. Gillan MM, Scofield RH, Harley JB (2002) Hashimoto’s thyroiditis presenting as bilateral knee arthropathy. J Okla State Med Assoc 95(5):323–325